Pakistan’s 1951 Census: State-Building in Post-Partition Sindh
Sarah Ansari
Royal Holloway, University of London, Egham, Surrey TW20 0EX, UK.
Keywords: Sindh; 1951 Census; citizenship; nation-building; state-building; Pakistan.
Abstract: This article explores the carrying out of Pakistan’s first (1951) census from the perspective of contemporary developments in the southern province of Sindh. Conducted against the backdrop of Partition-related migration to and from the province, this attempt at population enumeration proved to be a mammoth bureaucratic undertaking on the part of the recently-created Pakistani state. The challenges that this exercise posed at the provincial level shed light on processes of attempted nation-building, as well as the centrality of population counting to the biopolitical management of citizenship, during a key period of transition in mid-twentieth century Sindh.
…the Pakistan Census Commissioner appealed to every citizen of Pakistan to see that he is enumerated properly and none of his family members are missed out, that nobody is enumerated more than once, and the enumerator is given a truthful answer to all the questions…1
From the perspective of the Pakistani province of Sindh, Partition’s main impact was demographic: as a territorial unit, unlike the Punjab or Bengal, the province’s recently- acquired borders remained intact, but the existing balance between the different groups making up its pre-1947 population was upset by the arrival of Muslims in huge numbers from other parts of the subcontinent and, in due course, the departure of many of its non-Muslims for what had become India.2 After World War II growing support for the Muslim League’s manifesto in Muslim-majority provinces, had translated into an increasing proportion of
Muslims in Sindh starting to welcome the idea of some kind of ‘Pakistan’ in the run-up to
1 The Pakistan Times (9 Feb. 1951). 2 Sarah Ansari, Life after Partition: Migration, Community and Strife in Sindh, 1947-1962 (Karachi: Oxford University Press, 2005), pp. 46-73. 2 independence, whether or not they understood this to involve the creation of a separate sovereign ‘homeland’ for South Asia’s Muslims. 3 But what took them, like other contemporaries across South Asia, by surprise was the sheer size of the population transfer that accompanied the process of partitioning British India into two independent successor states. In this context, earlier sensitivities regarding the presence of ‘outsiders’ were revived, producing contradictions in the immediate post-Partition policies of Sindhi administrations: on the one hand, local Muslim politicians swung between welcoming incomers and cautioning against too many of them arriving; on the other, sometimes they urged non-Muslims to stay, while, on other occasions, facilitating (or even encouraging) the latter to leave.4
Sindhi ambivalence over what ‘Pakistan’ promised predated independence. By early
August 1947, despite support for the creation of a separate Muslim state in the Sindh
Legislative Assembly, there was nervousness about what the future held for the province itself. In the words of a contemporary observer:
Many persons in Sind [sic] are beginning to wonder as to whether the establishment of the Pakistan Government in Karachi may not result in the province of Sind [sic] becoming swamped by an inflow of persons from other parts of the new Dominion, which will result in this province being reduced to an insignificant position.5
3 For recent different approaches towards understanding the political thinking that propelled the so-called Pakistan Movement, see Faisal Devji, Muslim Zion: Pakistan as a Political Idea (London: Hurst and Co., 2013) and Venkat Dhulipala, Creating a New Medina: State Power, Islam, and the Quest for Pakistan in Late Colonial North India (Cambridge: Cambridge University Press, 2015). Both studies, however, focus primarily on the priorities and strategies of Muslims operating in parts of the subcontinent where they constituted a local minority. There has been far less interest in late directed towards accounting for how and why Muslims in so-called ‘Muslim-majority’ provinces came to support the Muslim League agenda, and so, in this respect, studies such as David Gilmartin’s ‘classic’ Empire and Islam: Punjab and the Making of Pakistan (Berkeley CA., University of California Press, 1988) remain key to our understanding, just as Ayesha Jalal’s The Sole Spokesman: Jinnah, the Muslim League, and the Demand for Pakistan (Cambridge: Cambridge University Press, 1985) underlined the importance of taking account of the contingent factors at work in Indian politics after 1939. 4 Ansari, Life after Partition, Chapter 4. 5 Transmission of a copy of an editorial in the Daily Gazette (5 Aug. 1947) entitled ‘Sind’s Place in Pakistan’, US Consulate Karachi, Despatch 77, 8 Aug. 1947, 845F.01/8-847, United States National Archives (henceforth USNA).
3
As this newspaper editorial suggested, the question of how independence would affect Sindh had already become a source of concern, with reports at the time (correctly it would appear) predicting that the issue was likely to cause serious friction in the longer-term. Not surprisingly perhaps, numerical anxieties—usually regarding people but about other matters as well—then proceeded to underpin political debates and disagreements regarding Sindh's place in post-Partition Pakistan.
It was in this fluid—political and demographic—context that Pakistan’s first national census was carried out in 1951. Conducted against the backdrop of Partition-related migration, it represented a mammoth bureaucratic undertaking on the part of the (still very new) Pakistani state, made tricky by the ongoing uncertainties generated by millions of people so recently on the move. At a provincial level, Sindh presented the census organizers with a particular set of tricky challenges linked to the demographic upheavals that the region was still experiencing. One specific difficulty that they faced was that census delimitation in the province had—in practice—to be based on very rough estimates of population, with potential inaccuracies linked to the continued arrival of migrants, their frequent changes in residence, and—importantly for local census officials—the lack of any proper record regarding the huge number of temporary dwellings that had been springing up since August 1947. In 1949 there were still an estimated 45,000 to 95,000 refugees in unofficial camps, while at least the same number again were to be found living in temporary shelters, all of which posed complications for the collection of census data.6 Likewise—to judge from contemporary reports—in the event it proved easier to count people located in rural parts of Sindh than in towns and cities: the disruptions to urban life caused by Partition-related migration presented more practical
6 Dawn (3 March 1949); ‘Status of Refugee Problem in West Pakistan’, Despatch 475, 9 Nov. 1949, 845F.48/11-949, USNA.
4 difficulties than in the (relatively more) settled countryside.7 The arrival of hundreds of thousands of muhajirs, together with over-spill migrants from other parts of Pakistan, posed enormous practical difficulties for administrators and enumerators at a time when the local bureaucracy lacked manpower and experience.
Without doubt, as the carrying-out of Pakistan’s 1951 census in Sindh highlighted, and as historians and other social scientists now acknowledge, Partition was an extended process that dragged on for many years in terms of the practical—nation-building—tasks that it posed.
Pakistan may have come into existence at the stroke of midnight on 14/15 August 1947, but as a state created from scratch it remained a ‘work in progress’ for long after its founding moment.8 Its first census, therefore, needs also to be viewed as a deliberate attempt at state- building, its effectiveness hinging not just on how systematically it categorised, enumerated and documented the population, but also on Pakistan’s recently-created citizens themselves sharing responsibility for the exercise working smoothly. Building on Foucault’s argument that “modern ‘governmentality’ takes population as its object”,9 many now view enumeration as “a critical modality of governmentality”: after all, “what kind of statistics are collected,
7 First Census of Pakistan 1951: Administrative Report, Part Two: The Reports of the Provincial Superintendents of Census Operations (Karachi: Government of Pakistan Press, 1953), Chapter 11 – Karachi (henceforth Karachi Report), p.123 8 See, among other studies on the afterlife of Partition, Yasmin Khan, The Great Partition: The Making of India and Pakistan (London: Yale University Press, 2007), Vazira Fazila-Yacoobali Zamindar, The Long Partition and the Making of Modern South Asia: Refugees, Boundaries, Histories (New York: Columbia University Press, 2007), T. Y. Tan and G. Kudaisya, The Aftermath of Partition in South Asia (London: Routledge, 2000); Yasmin Khan, ‘The Ending of an Empire: From Imagined Communities to Nation States in India and Pakistan’, in The Iconography of Independence: Freedoms at Midnight, eds. Robert Holland, Susan Williams and Terry Barringer (London and New York: Routledge, 2010), pp. 47-56. For a contemporary assessment of the challenges faced by Pakistan immediately after Partition, see Richard Symonds, ‘State Making in Pakistan’, Far Eastern Survey, Vol. 19, no. 5 (8 March 1950), pp. 45-50. 9 Bruce Curtis, ‘Foucault on Governmentality and Population: The Impossible Discovery’, The Canadian Journal of Sociology / Cahiers canadiens de sociologie, Vol. 27, no. 4 (Autumn 2002), p. 505.
5 who collects them, and how they are used all affect the regulation of populations, techniques of accountability, and the formation of group identities”.10
From this perspective, the decision to hold a census provided the Pakistani authorities at provincial as well as federal level with a valuable opportunity to project the role of the state as well as to communicate the responsibilities that they associated with being its citizens.
Benedict Anderson in his influential Imagined Communities argued that censuses enable the state, or those who control and direct it, to view and quantify society: for him, they were the way, for instance, in which a colonial state imagined and hence controlled its dominion.11
But, it could also be maintained, that from the perspective of the people being surveyed, and all the more so where illiteracy is high, the act of conducting a census also brings citizens face-to-face with the state as the latter’s representatives engage in collecting and recording the required information. In the words of an earlier (1941) British census official, “the cardinal fact [is] that it takes two to make a census, the enumerator and the citizen, and that of these two the role of the latter is the more fundamental and vital. The enumerator broadly is a scribe: in any census it is the citizen's answers which are sought and are tabulated”.12 The reality proved to be very similar in post-Partition Sindh. With extremely restricted literacy
(the census itself eventually revealed a national rate in the vicinity of 13 percent), the collection of census data relied heavily on volunteer enumerators—frequently praised for their patriotic service13—in effect taking the state along with their questionnaires directly, and what was in most cases for the first time, into people’s homes or living spaces, in the process
10 Akhil Gupta, Red Tape: Bureaucracy, Structural Violence and Poverty in India (Durham and London: Duke University Press, 2012), p. 262. 11 Benedict Anderson, Imagined Communities: Reflections on the Origins and Spread of Nationalism (London: Verso, rev. ed., 1991), pp. 163-85. 12 Note from Census Commissioner, Census of India, 1941 [http://censusindia.gov.in/Census_And_You/old_report/Census_1941.aspx, accessed 8 May 2016]. 13 Dawn (11 Feb. 1951).
6 spreading the word about what the state now represented, or at least claimed that it did.14 As others have suggested in other contexts, one way that people learn to recognise themselves as members of a nation-state is by participating in the ritual of a national census.15
Numerical anxieties in Sindh
In the decades preceding independence and partition, like other parts of South Asia, local
Sindhi politics had acquired a distinct numerical twist with numbers assuming a growing political significance. Even without the drawing power of many major urban centres (the main exceptions being Karachi, Hyderabad and Sukkur), Sindh’s population grew rapidly in the early twentieth century, particular in the smaller towns and districts affected by the expansion of irrigation facilities.16 By the time of the penultimate colonial census in 1931, districts in Sindh were showing a faster rate of growth than in the average district elsewhere in Bombay Presidency, even though the increase in irrigation flowing from the completion of the mammoth Sukkur Barrage did not begin until 1932: between 1921 and 1931 Sindh’s overall population density per square mile increased by nearly 20 percent, as compared with around 13 percent in other Bombay Presidency districts.17
But, while the creation of a separate province in 1936 hinged in large part on financial arguments with the economic effect of the Sukkur Barrage a key point in its favour,18 the
14 Khwaja Nazimuddin, Pakistan’s Governor-General by this time, was the first person to be counted, followed by the people living in Prime Minister Liaquat Ali Khan’s household, see The Pakistan Times (9 Feb. 1951) . 15 Vicente Berdayes, ‘Official Statistics as Curriculum: Biopolitics and the United States “Census in Schools” Program’, Review of Education, Pedagogy and Cultural Studies, Vol. 30, no. 2 (April 2008), pp. 85-114. 16 Radhakamal Mukerjee (ed.), Economic Problems of Modern India (London: Macmillan, 1939), Vol. 1, p. 14. See also Census of India, 1931, Vol. 8 (Bombay Pres.), pt. 1 (Delhi: Manager of Publications, 1933), p. 29, cited in Kingsley Davis, The Population of India and Pakistan (Princeton NJ: Princeton University Press, 1951), p. 120. 17 For data on interwar population growth in Sindh, see Government of India, Census of India 1941, vol. XII (New Delhi: Government Press, 1942). 18 Daniel Haines, Building the Empire, Building the Nation: Development, Legitimacy, and Hydro- politics in Sind, 1919-1969 (Karachi: Oxford University Press, 2013).
7 separation campaign also acquired an ominously communal dimension thanks to the statistical fact that local Muslims outnumbered local non-Muslims by a ratio of roughly three to one.19
Thus, demographic developments in Sindh combined with resentment at being administered by distant Bombay made at least some Sindhi Muslims increasingly sensitive to the realities of Bombay’s Hindu-majority and, in turn, this complicated their relations with non-Muslims closer to home.20 ‘Sind for Sindhis’ became a slogan of the campaign launched by local (in the main Muslim) politicians to secure their own ‘unattached’ province. Concern at unemployment was translated itself into calls for recruitment to government service, from peon-level upwards, to be reserved for ‘Sindhis and Sindhis alone’. 21 And following separation itself, the presence of non-Sindhis in the province remained under close scrutiny in the provincial legislative assembly, with MLAs issuing periodic demands to limit employment opportunities open to so-called ‘outsiders’, in particular Punjabis whatever their religious affiliation. Indeed, as I have discussed elsewhere, Sindhi politics during the decade before independence were greatly affected by the long-term consequences of colonial development policies that had encouraged the arrival of Punjabi settlers to work the lands released for production by irrigation schemes, and which helps explain Sindhi sensitivities towards the wave of newcomers generated by Partition. In addition, the increased commercialisation of agriculture taking place in the 1930s accompanied as it was by the growing importance of small towns—which became key economic hubs—contributed to perceptions among at least some Sindhi Muslims that it was non-Muslims with extra- territorial connections who were benefitting most from the new economic opportunities on
19 According to the 1931 Census, this was the overall ratio of Muslims to Hindus, but in urban areas Hindus outnumbered Muslims by 29 percent. See Census of India, 1931, Vol. 8 (Delhi: Manager of Publications, 1933), p. 50. 20 For an overview of the movement to secure a separate province of Sindh, see Allen Jones, Politics in Sindh 1907-1940: Muslim Identity and the Demand for Pakistan (Karachi: Oxford University Press, 2001), pp. 16-28. 21 For instance, The Sind Legislative Assembly Debates, Official Report, Vol. 2, no. 8 (11 Aug. 1937), pp. 14-17.
8 offer. All this helped to fuel, and helps to explain, post-1947 Sindhi anxieties about the place of muhajirs in the province.22
After Partition, the continuing political significance of numbers in the province was symbolised by Sindh’s loss of Karachi in 1948, when financial number-crunching together with its newly acquired status as the country’s capital played an important role in the decision to convert the city into federal territory. Accordingly in July 1948 the Karachi Municipal
Area, together with the civil cantonment, other neighbouring cantonments and 44 surrounding villages (which had previously formed part of the district of Karachi) were amalgamated to form a centrally-administered enclave, leaving what remained of it, renamed Thatta District, in Sindh. Power and responsibilities equating to those of a full-fledged province were transferred to an administrator. As one observer commented, Sindh had been “beheaded”.23
To a great extent, Sindh’s financial prospects—as a small province with limited resources— directly depended on whether or not the federal authorities adequately compensated it for the loss of Karachi revenues, but in the view of a large number of provincial politicians this matter was not satisfactorily addressed. By 1949, Sindh’s budget had reached a deficit of nearly ten million rupees, prompting its Finance Minister to observe in his 1949 budget speech: “we are reduced to the position of a very small province with very limited resources”.24 This deteriorating financial position meant that expenditure had to be cut—out of 88 development schemes, a quarter were abandoned and the remainder subjected to drastic review and revision.25
Numerical anxieties likewise entered the provincial political equation when during the same year—1948—there was intense disagreement over whether Sindh should absorb large numbers of overspill refugees from neighbouring Punjab. By the start of that year nearly a
22 Ansari, Life after Partition, pp. 30-9. 23 Dawn (10 Feb. 1948). 24 Quoted in Commerce (20 Aug. 1949), Despatch 314, 27 Aug. 1949, 845F.50/8-2749, USNA. 25 ‘Provincial budgets of Pakistan’, Despatch 106, 1 April 1949, 845F.51/4-149, USNA.
9 million displaced people remained in Punjabi refugee camps, including around 250,000 living in the open air. To the federal government, relocating them southwards seemed like an obvious solution. But Sindh’s political leaders rebuffed this idea. Instead they insisted on a limit of 100,000—in the view of the then provincial Chief Minister MA Khuhro, Sindh could not absorb more than this number: in his words, “for every one Hindu that has left, two
Muslims have come in”, and he rejected any further increase on the grounds of insufficient resource to cope with the influx. 26 In response, the Federal Minister for Refugee
Rehabilitation Ghazanfar Ali Khan argued that no province could impose any limit on its intake of refugees. Condemning what he described as the narrow spirit being exhibited by the
Sindh authorities, he warned against the “virus of provincialism” that, for him, contradicted both the teachings of Islam and the principles on which Pakistan had been fought for and won.
If this “virus” went unchecked, he cautioned, it would “destroy the very foundations of our newly born state”.27 In the end, more than 200,000 refugees arrived. From early September onwards, they travelled at the rate of roughly 5,000 per day, 2,500 per train. Meanwhile refugees from India, together with migrants from elsewhere in Pakistan, kept on arriving in the province; by 1949, more than 700,000 incomers had arrived there, with most heading for
Karachi and many of the remainder settling in other urban centres in Sindh, a pattern that continued well into the 1950s.28
Organising a National Census
National censuses were nothing new in former British India. As others have shown, the colonial authorities had made attempts to gather British India’s vital statistics from the early
26 Dawn (18 Jan. 1948). 27 Dawn (24 Feb. 1948). 28 See Ansari, Life after Partition, Chapter 5.
10 nineteenth century onwards.29 Initially these had been periodic number-counting operations carried out usually for specific purposes and for limited areas, including annual provincial reports on medical relief that covered epidemic diseases (such as Presidency Board reports as early as the great epidemic of 1817), hospitals (beginning in 1851), and vaccination
(beginning in 1861). Famine too generated statistical reports, as did other matters concerning mortality such as in relation to the army, jails and certain key cities that went back to the first half of the nineteenth century. In 1871, the first general census for the whole of British India at a given date was carried out. India under British control became a ‘happy hunting ground
[for] official investigating bodies’, which in turn developed a reputation for producing suitably voluminous official reports on a vast range of subjects.30 .
Censuses in British India, however, had also acquired a reputation for having to overcome a set of especially challenging practical obstacles. As one commentator put it,
“Imagine a massive, diversified subcontinent with hundreds of millions of people nearly all of whom are illiterate, most of them rural and some isolated in jungles or mountains, some harbouring superstitions inimical to census cooperation, some split by political and religious rifts … imagine all this and the difficulty of taking a census becomes apparent”. Under these circumstances, the same source continued, it was “no wonder the hospital rate for census officials [had] been high”: the 1941 effort had reportedly sent four provincial superintendents
29 For discussion on the census and other information-collecting exercises in British India, see, among other publications, N. Gerald Barker (ed.), The Census in British India: New Perspectives (New Delhi: Manohor, 1981); Crispin Bates, ‘Race, caste and tribe in Central India: The early origins of Indian anthropometry’, in The Concept of Race in South Asia, ed. Peter Robb (New Delhi: Oxford University Press, 1995), pp. 219-59; Bernard Cohn, ‘The Census, Social Structure and Objectification in South Asia’, in An Anthropologist among the Historians and Other Essays, ed. Bernard Cohn (Delhi: Oxford University Press, 1987), pp. 224-54; Ronald Inden, Imagining India (Oxford: Basil Blackwell, 1990); Shriram Maheshwari, The Census Administration under the Raj and After (New Delhi: Concept Publishing, 1996); and Adam Bowles, ‘An Introduction to the Indian Census’, [https://arrow.latrobe.edu.au/store/3/4/5/5/2/public/census.htm, accessed 8 May 2016]. 30 Davis, Population of India and Pakistan, pp. 5-6.
11 to hospital, leaving others “in a weakened condition” while “the [Census] Commissioner himself was ill eleven months out of the year”.31
Political complications could also on occasion intervene. The 1931 Census, for instance, had coincided with and been confronted by Congress’s Non-Cooperation
Movement, which had designated 11 January that year as ‘Census Boycott Sunday’.32 But, despite such complications, census officials in British India had importantly tended to regard their duties as transcending the mere gathering and reporting of statistics. As well as interpreting census results, these officers also discussed their methods frankly, exposed their difficulties, shared their misgivings and expressed their plans and hopes: they did not hide
“behind a wall of infallibility”.33 By 1941, it had been recognised that the old-style omnibus report was now out-of-date, and so the then Chief Commissioner MWM Yeats proposed cutting back the scope of census volumes to include only such content as was deemed necessary for interpreting, and understanding, the tables:
The great old Reports of the Indian Census—some of them classics for all time— were written with the view to lead and guide all thinking in India in the matter of analysing and correlating the immense mass of figures which at every census was collected…but now everywhere we find a group of scholars who are competent to deal with the measurement of social phenomena, as in western countries, and the role of the counter of people has to approximate itself to that of his opposite number in England or America,—the humbler role of just collecting statistical data, doing as little as possible of writing and leaving the figures to speak for themselves.34
Pakistan’s political leaders, however, were not prepared just to let the numbers do the talking. While Pakistan’s first exercise in nation-wide number-counting was influenced—at least to some extent—by the legacy of earlier censuses conducted under British rule, it was also affected by the challenges created by the after-effects of Partition and the broader
31 Ibid. 32 Ibid., pp. 4-5. 33 Ibid. 34 Government of India, Census of India, 1941, Vol. 1 (New Delhi: Government Press, 1942), p. 2.
12 pressures associated with creating a new state. From the moment of Pakistan’s creation,
Muhammad Ali Jinnah, Liaquat Ali Khan, and other politicians in different parts of the country hammered home messages about the need for the state—its representatives and its people—to pull to together on behalf of the nation. Shortly after independence, Jinnah used the occasion of Pakistan’s first Eid celebration in August 1947 to remind its citizens that
No doubt we have achieved Pakistan, but that is only yet the beginning of an end. Great responsibilities have come to us, and equally great should be our determination and endeavour to discharge them, and the fulfilment thereof will demand of us efforts and sacrifices in the cause no less for construction and building of our nation than what was required for the achievement of the cherished goal of Pakistan. The time for real solid work has now arrived, and I have no doubt in my mind that the Muslim genius will put its shoulder to the wheel and conquer all obstacles in our way on the road, which may appear uphill.35
But despite such stirring speeches, the Pakistani authorities found themselves constrained on a practical administrative level by a shortfall in the number of experienced officials needed to manage day-to-day operations, made all the trickier by the lingering fallout from the demographic and economic dislocation produced by Partition.36 On the one hand, there was a desire and need to distance the state from its colonial predecessor; on the other, the independent state in practice often did not seem so very different—certainly from the perspective of many ordinary people engaging with it on a quotidian basis—from what had existed in the years leading up to 1947. Official rhetoric, almost as soon as the moment of independence had come and gone, repeatedly laid stress on the fact that much had changed, but in many ways—from ration cards that harked back to wartime rationing, to the enduring blight of corruption and the continued reliance on politicians who lacked a proper democratic mandate—life in Pakistan in many ways bore an unsettling resemblance to that of pre-
35 Jinnah, Eid Speech, 18 Aug. 1947 [http://m-a-jinnah.blogspot.co.uk/2010/03/first-eid-in- pakistan-18th-aug-1947.html, accessed 8 May 2016]. 36 William Gould, Taylor C. Sherman and Sarah Ansari, ‘The Flux of the Matter: loyalty, corruption and the “everyday state” in the post-partition government services of India and Pakistan’, Past & Present, Vol. 219, no. 1 (2013), pp. 237-79.
13 independence days. Popular expectations were high thanks to the shared euphoria of independence, but by the early 1950s many people living in Pakistan were continuing to express disappointment at the everyday realities of a state that was still very much in-the- making.37
From the outset, therefore, Pakistan’s first census was presented to the public as a hugely crucial nation-building test. 38 For the Federal Minster of the Interior Khwaja
Shahabuddin, speaking on Radio Pakistan in February 1951, its success hinged on teamwork between the public and the mass of volunteers who were to collect the necessary data:
The gathering of all this information is a work that is being undertaken by a large body of unpaid workers.…To all these, Pakistan owes a debt of gratitude for their selfless labours [sic]…Throughout most of Pakistan the month of February will see them very busily employed. When these gentlemen visit our homes…we must treat them with courtesy and respect as men who are doing a vitally important job which will be of the greatest help to everybody. 39
In addition, the Minister went to great lengths to refute popular misconceptions that any information provided would be used to disadvantage individuals or that the census would arrive at a preconceived result: “We are out to get at the truth whatever it may be…and I am sure that the people of Pakistan and particularly the census officers will ensure that everyone is included once and only once, and that slips are not made out for non-existent people”.40 In similar fashion, one correspondent in Dawn (who was keen to emphasise that refugees in
Karachi could be “taken into full confidence and their full cooperation secured”) reminded other readers that “After all, [a] Census is held only once in a decade, and is a national service” and, accordingly, since it was clear that “this inescapable duty should not be thrust
37 Sarah Ansari, ‘Everyday Expectations of the State during Pakistan's Early Years: Letters to the Editor, Dawn (Karachi), 1951-53’, Modern Asian Studies, Vol. 45, no. 1 (2011), pp. 159-78. 38 Dawn (20 June 1950). 39 The Pakistan Times (10 Feb. 1951). 40 Ibid.
14 upon unwilling workers”, he urged the Census authorities to take “steps to enlightened the general public about the general scheme and principles of Census operations”.41
But, like a lot of things that happened before and after August 1947, the first Pakistan census was organised in a hurry. As its Commissioner EH Slade later commented in his official report, a “lack of time and the necessity for speed” created the “all-pervading atmosphere of the Census of 1951”.42 Thanks to disruptions and manpower shortages in the administrative services caused by the creation of two separate states at independence, a
Census Act (Census Act (No. VII) of 1950) was not placed before Pakistan’s Constituent
Legislative Assembly until late December 1949 when it was then rapidly processed, receiving the assent of the Governor General (by then Khwaja Nazimuddin) the following month. On 1
May 1950 a Ministry of the Interior press notice announced that the count would take place in
February the following year. At the same, the announcement also explained that “along with the traditional counting of heads, various useful data regarding religion, nationality, age, birth place, domicile, mother tongue, marital status, literacy, economic characteristics, agriculture, industries, and the migration of population’ would be ‘collected and tabulated for future use by the State and the public”.43
As no suitably-experienced officer was to be found in Pakistan’s already over- stretched central bureaucracy, 44 the post of Census Commissioner was advertised
41 Dawn (10 June 1950). 42 First Census of Pakistan: Administrative Report, Part 1: The Census Commissioner’s Report (Karachi: Government of Pakistan Press, 1953), p. 2 (hereafter CCR). 43 Press Note, 1 May 1950, Ministry of Interior (Home Division), Government of Pakistan, MSS Eur F 158/628, British Library (hereafter BL). 44 As early as April 1948 the UK High Commissioner in Pakistan reported that a census was to be held there in 1951 and that initial steps were being taken to prepare for this, including inviting a census expert from the UK. See General Register Office, Somerset House, London, to Commonwealth Relations Office, 7 May 1948, in Pakistan Census, IOR E (B) 4949/48 BL. In the event it took two further years before this post was advertised and filled.
15 internationally, leading to Slade's appointment and arrival in Karachi in June 1950.45 This gave him just over six months in which to put the necessary mechanisms in place.46 In the interim, steps were taken to secure the nomination of suitable officers as Superintendents of
Census Operations in each province, negotiations started with central ministries regarding the information that was to be sought through the census, and a census budget drawn up and approved by the Ministry of Finance. Despite being given such a short space of time in which to deliver on his promise of carrying out a census by early 1951, Slade remained confident that his recent experience of organising a similar exercise under equally rushed circumstances in post-war Germany would ensure success, with “Pakistani enthusiasm and patriotism
[substituting for] the efficiency of German officials and the compelling power of Military
Government”.47 In July 1950 a conference in Karachi brought together representatives from all the relevant ministries to settle on the scope and general method of census operation to be deployed. However, it did not take long for the officials involved to recognise that if the count was to take place as planned at the beginning of 1951, there was not sufficient opportunity to make any serious changes to the organisational methods that had been used in the previous 1941 census. As a result, they agreed that “the main lines of the operation [had to be] kept as simple as possible”.48
Slade in his final census report eventually published in 1953 cautioned strongly against any future repetition of such a tight turn-around. With the benefit of hindsight, he believed that this constraint had made the whole operation far more stressful than it had needed to be. All the same, it is important to note that—despite the practical challenges posed by Independence and Partition—censuses during the British period had also tended to be held
45 Slade’s credentials included being a Fellow of the UK’s Royal Statistical Society, and also a Fellow of the Association of Incorporated Statisticians (London). 46 CCR, p. 2. 47 Ibid. 48 CCR, p. 3.
16 at relatively short notice. In turn this had prompted earlier census commissioners to criticise strongly what they had informally termed as the ‘Phoenix System’—namely, officials being appointed only a couple of years ahead of a census, carrying out the enumeration, tabulating the results and then disappearing along with the valuable experience that they had acquired in the process—which meant that each successive census operation under colonial rule had been unable to learn full lessons from the previous one. 49 But, in practice, there was little opportunity in the run-up to Pakistan’s 1951 census to study what had and had not worked in the past, had Slade and team wished to do so. Circumstances did not allow this to happen. For instance, Slade later confessed that when he had tried to track down any available records of how earlier censuses had been conducted in the territory that now comprised Pakistan, he had come across two godowns in the Secretariat compound in Lahore which, though they contained documents from the 1941 Census, had been locked up for eight years, serving “no better purpose than the nourishment of white ants”:
The only things of use were a damp and damaged but still complete set of Census Reports for 1931 for the Punjab. These have been very difficult to obtain since partition [and so were very welcome] ... But the rest of the treasure...had by the mere effluxion [sic] of time become perfectly useless. We retained a few samples and agreed that the rest might be destroyed.50
In the event, some 200,000 enumerators, 16,000 circle supervisors, 1,770 charge supervisors, and 88 district census officers took part in what proved to be a massive state undertaking culminating in Census Night of 28 February-1 March. Controlling the field operation were the offices of the Census Commissioner, five Provincial Superintendents, and the Chief Census Officer for the Federal Capital Area that had been so recently, and controversially, separated from the rest of the province of Sindh. Interestingly, these officers were deliberately chosen from among senior provincial civil servants:
49 CCR, p. 4. 50 Ibid.
17
For the most part they were nearing the end of their active life.… None of them had any census experience to speak of and none were trained or had even read much about statistics. Their acquaintance with mathematics was for the most part a memory of their student days. Each of them, however, was an experienced administrator with a most intimate knowledge of his province, its people and its Governmental organisation. They had many friends in all places and at levels in the Provincial civil service and among the public. For the purposes of organising the enumeration, therefore, they were perfect.51
In addition, around a quarter of a million volunteers—comprising mainly students, government employees, school teachers, Muslim League workers and Muslim League
National Guards—were deployed by the Pakistani authorities, who divided the country into roughly 160,000 blocks, with ten blocks making up a circle, and every ten circles a charge. It was the enumerator’s duty to visit every household in his area and obtain answers to a set of
16 questions in respect of “every individual citizen”, covering
… all the important demographic and economic characteristic, individual age, sex and marital status, nationality, religion, languages spoken, written and read, education and occupation. Special questions were asked from the persons, known as Muhajirs, who had come into Pakistan leaving their home in other parts of the sub-continent of India as a result of Partition. The agricultural and economic status of members of the civilian labour force were recorded, and certain questions asked to married women with a view to obtaining fertility data.52
And once the period of enumeration itself was concluded, official reports took public pride how far the work involved had been “carried out with enthusiasm and a great measure of patriotic self-sacrifice”. Praise was equally heaped on “the [apparently] friendly and willing cooperation of the public” together with the “tactful and pleasant attitude adopted by the enumerators”:53 as a later report claimed, “one only [had] to examine [the enumeration slips] to realise that their preparation was a labour of love” for the country.54
51 CCR, p. 7. 52 Pakistan Information Department, Pakistan Day: Fourth Anniversary Series, Article No. 27/51, ‘The First Census of Pakistan – 1951’, MSS Eur F 158/628, BL. 53 Press Note, E No. 1061, 20 March 1951, Press Information Department, Government of Pakistan, MSS Eur F 158/628, BL. 54 Ibid.
18
Census Operations in Sindh
But a closer look at developments in Sindh and the recently-separated federal capital of
Karachi suggests that the reality on the ground could prove very different to the positive officially-endorsed gloss put on the census operation before, during and after it was carried out. The problems caused by the tightness of the census timetable together with insufficient numbers of experienced officials and willing volunteers proved especially acute in Sindh, and, hence, official reports singled out the province in general and the Federal Capital Area in particular as having presented a distinctive set of difficulties for those charged with carrying it out.
Unlike the 1941 census when local planning in Sindh had started more than two years ahead of the operation (HT Lambrick, who had headed up the provincial census authorities a decade earlier, had his staff in place from October 1939 onwards), the office of Muhammad
Hashim, the Provincial Superintendent of the Census for Sind and Khairpur State, was given only nine months—a “dangerously short period”—in which to make the necessary preparations. Hashim, himself only appointed in May 1950 (and before Slade himself had arrived in Pakistan), was forced to select his staff without knowing even what grade of pay they were to be offered. It took until the end of November 1950—three months before the census was scheduled—for final clarification on this matter to be received, by which time both his first-choice Head Clerk and his Camp Clerk had—within five weeks of their appointment—rather inconveniently disappeared to study for the necessary qualifying examinations. Moreover, while the Head Clerk was said to have “worked keenly in the hope that he was to be raised to a Gazetted Post and his pay doubled,… when he found out that the
Census gave an increase of only Rs. 35/- over his basic pay, he lost interest and proceeded on
19 leave on medical certificate”.55 The understandable stress involved in finding and holding on to an adequate number of trained staff who did not believe that they were being paid sufficiently accentuated the difficulties caused by the lack of time, and for contemporary observers “it [was] a wonder that the Enumeration [in Sindh] was a success at all”. 56
Relatively-speaking, therefore, with the provincial civil service having suffered a double disruption—caused first by Partition and then by the separation of Karachi that syphoned off potential men—the census operation here arguably faced a greater challenge thanks to the lingering impact of Partition than did other parts of Pakistan.
Based in one half of the Deputy Collector of Hala’s office, Hashim was also initially given responsibility for the Federal Capital Area alongside the eight districts that then comprised the province and Khairpur State. Karachi, however, was in due course given its own separate census apparatus. But thanks to the early confusion regarding who or what would be responsible for the census work in the city, its eventual own chief census officer
AM Jafri only assumed his duties at the end of August 1950. Throughout the process Jafri’s staff remained minimal, comprising three clerks of different grades, a typist and two peons.57
Much of the reason for this delay and resulting confusion was connected to problems linked to how to disengage or separate the newly-created FCA from its former provincial moorings. As a result, the decision to treat the city and its surroundings as a separate census province with its own Chief Census Officer was only taken once the other provincial superintendents of census had already started their work. The delay was also partly due to problems caused by the Karachi Municipal Corporation's assumption that, as during the previous 1941 Census when the then authorities had treated Karachi as a special case, the
55 First Census of Pakistan 1951: Administrative Report, Part Two: The Reports of the Provincial Superintendents of Census Operations (Karachi: Government of Pakistan Press, 1953), Chapter 14 - Sind (hereafter Sind Report), p.196. 56 Ibid., p. 188. Hashim himself also had to be replaced on medical grounds before the census work was completed. 57 CCR, p. 29.
20
Pakistan government would meet 50 percent of the expense involved. But as the 1950 Census
Act stipulated, municipal corporations were now financially responsibility for census work within their administrative areas. In the case of Karachi, it seems to have been only once
RAF Howroyd became Municipal Commissioner late in 1950 that Jafri and his team finally secured the all-important cooperation from Corporation employees.58
Karachi for census purposes was divided into two districts, with the Collector of
Karachi given responsibility for the first that included cantonments and villages, while the second comprising the central parts of the city was placed under the direction of the
Corporation’s chief officer. The number of census sub-divisions or ‘charges’, however, fell far short of what had originally been envisaged. The Civil Cantonment, for instance, with an area of about four square miles and a population by now of over 100,000, remained a single unit, despite Jafri's efforts to sub-divide it into more manageable component parts. On the other hand, ‘District No. 2’ was split into 44 charges, including the Port Trust, the Railway and the Customs areas as separate units. What further complicated the operation was that the census in Karachi had originally been planned with existing wards as the basis for charges, but this proved impossible to coordinate on account of the huge increase in the city’s population since 1947. Because municipal wards themselves varied considerably in size, and contained fluctuating numbers of inhabitants, the charges were not uniform and this created particular problems as far as accurate data collection was concerned.59
Elsewhere in Sindh, senior local officials—such as the Deputy Commissioner of
Upper Sind Frontier (USF), various District Collectors and the Chief Officer of Hyderabad
Municipality—assumed responsibility for local census operations. Districts were then sub- divided into charges, mainly on the basis on existing revenue talukas, mahals and municipal areas. At this level, the key official was the mukhtiarkhar in rural areas and the chief officer
58 Ibid. 59 Karachi Report, p. 123.
21 or secretary for towns. Charges were in turn split into ‘circles’, each placed under the control of a supervisor and split into blocks comprising—at least on paper—not more than 150 houses in rural areas and fewer than this where possible in urban settlements. An average of
300-500 persons per block was adjudged to be the right number for a single enumerator to handle conveniently and efficiently. In practice, the number of people in most blocks turned out to be much higher, averaging nearly 650 in the province as a whole and reaching around
900 in the cities of Hyderabad and Khairpur. But the census operation in the province was also marked by recruitment problems from the outset, first when it came to raising the necessary number of charge superintendents and then, once the central authorities had made a number of their own officers available for this duty, in securing local supervisors and enumerators. The initial plan had been to make very wide use of non-officials, but it proved very difficult to obtain enough of them on the necessary voluntary basis. Frequently, as census officials reported, an early show of enthusiasm faded fast once volunteers learned of just how much hard work was going to be involved. On top of this reluctance, many would- be volunteers withdrew their offers of help when they found that they would not receive any payment for their efforts.60 In parts of Karachi, for instance, this triggered complaints that
“neither public men nor officials were willing to devote part of their time to this national duty”: out of 44 charge officers, the fact that only three were non-officials represented a “sad commentary” on the national spirit of public leaders in this part of the country.61 As the official census report for Karachi later complained:
The Central Government and the Government of Sind nominated certain officers for this honorary work but many of these managed to avoid it on plea or another with the result that the Charge Superintendents were not fully appointed until late in January 1951, with enumeration due to start in less than a month.62
60 CCR, p. 31. 61 Karachi Report, p. 123. 62 Ibid.
22
The difficulties encountered by Hashim and Jafri as far as attracting a sufficient number of charge superintendents proved to be nothing as compared with the monumental problem they faced when it came to recruiting the required quantity of volunteer enumerators and supervisors. The initial 3,000 or so enumerators who initially offered their services in
Karachi dwindled drastically once it was discovered that all they would receive for their involvement was out-of-pocket expenses and a sanad (certificate). With only ten days left before the enumeration started, two charge superintendents in the city complained that they had still not secured a single supervisor or enumerator. Despite the Collector of Customs, the
Port Trust Chairman, the naval authorities and the Muslim League National Guard all coming
“to the rescue at the eleventh hour”, a shortfall remained, which even the efforts of “devoted” helpers (including “an enthusiastic group of ladies who though not themselves holders of any census office [enumerators had to be male] rendered great help to the organisation”) could not disguise.63
Throughout the planning stage, directives generally trickled their way down to the officials and helpers involved. In July 1950 a conference of district census officers took place in Hyderabad at which important points of procedure were communicated by Slade to his team. These guidelines were then passed on to charge superintendents, who in turn took responsibility for training supervisors and enumerators. Thereafter it became the duty of enumerators to arrange local meetings to explain to people in their blocks the object and scope of the census and to appeal for their cooperation. Meanwhile, the authorities launched a massive publicity campaign in newspapers, on the radio and with the use of posters in public places in villages and towns alike. In the city of Hyderabad, for instance, vans with loud speakers toured the streets while exhibits in local cinemas promoted the same instructions.
63 Ibid., pp. 123-4
23
Radio Pakistan broadcast talks on why the census was so important to the country, which included Hashim addressing listeners in Sindhi on how to fill in the census slips.64
One of the enumerators’ preliminary duties was to prepare their blocks in readiness for the collection of census data. As well as being supplied with house-list forms in both Sindhi and Urdu, they were given the necessary red and white paint with which daub numbers on the houses making up their respective blocks. While it took until mid-September 1950 before the house numbering had started across the whole province, in most places the process encountered few problems, one exception being the town of Nawabshah where it was reported that (but not explained why) some of these numbers were rubbed off and so needed to be repainted. However, in the main, completing this numbering task together with house listing took longer than anticipated. Only in Dadu district was it achieved by the deadline of mid-
December. Elsewhere (apart from USF and Larkana where it was finished by the end of
December) it was not until January (less than a month before the start of the census period at the end of February) that these census preparations were completed (and even then lingering trouble with local Hurs in the desert portions of Khipro taluka of Thar Parker district meant that house listing there had to be carried out at the same time as enumeration a month ahead of the official census period 65 ). In Karachi, the purpose behind the household listing operation generated disquiet in view of legal complications associated with the ongoing scramble for living accommodation. Under these circumstances, Slade was required to issue the following reassurance:
… telling the truth in the household list helps to make the population census valuable to Pakistan but it has no legal use and gives no information that could lead to action in individual cases. No one need feel that the list could be used as a basis for any action by housing authorities.66
64 Sind Report, p.195. 65 For background on the Hur ‘problem’ in Sindh, see Sarah Ansari, Sufi Saints and State Power: The Pirs of Sind, 1943-1947 (Cambridge: Cambridge University Press, 1992), Chapter 6. 66 Dawn (11 Jan. 1951).
24
With accommodation still such a pressing issue in Karachi, and the city’s population continuing to swell, people there were clearly sensitive about how information about their living arrangements might be put to use by the authorities. While large numbers of people remained unhoused in permanent dwellings, and as Zamindar’s work on the “economics of displacement’ in Karachi in particular has so persuasively demonstrated, individual refugees as well as groups representing them were quick to criticise what they viewed as corrupt practices as far as the allocation of housing was concerned. They were also suspicious about attempts by the authorities to re-distribute evacuee property, often assuming that the system was being manipulated in favour of those with connections in the right places rather than on the basis of need.67
Unlike the 1941 census, when household lists in Sindh had been prepared first, followed by the numbering of houses and counting of household inhabitants over a three- month stretch, these previously separate steps took place simultaneously in the run-up to the
1951 count.68 And whereas the wartime census had aimed at enumerating people at their normal place of residence (that is, where they had lived for the previous six months), now the focus was simply on who happened to be in a household at the moment when the census was conducted. With so much population flux in the period since Partition (and with continued movement into and out of Pakistani territory) any other approach was rejected on the grounds that it would risk overcomplicating the results.
The actual census period in Sindh also varied markedly between urban and rural areas.
In the former, enumeration took place between 9 February and 1 March (Census Night), just as it did in most other parts of the country. In Hyderabad city, for instance, the count was completed according to schedule without any apparent difficulty. But in the countryside, this
67 Zamindar, The Long Partition, Chapter Four ‘Economies of Displacement’. See also Ansari, ‘Everyday Expectations’ for more discussion of the day-to-day frustration of ordinary people in Karachi at the failure of state in the early 1950s to address their needs. 68 Sind Report, p. 198.
25 period was initially extended to start from 1 January but, as it turned out, this could not happen as the necessary arrangements had not been fully completed by this earlier date. In some parts of the province—USF, Larkana, Thar Parkar and Nawabshah districts— enumerators encountered resistance when it came to giving information about women, for instance on matters concerning fertility or births, with the result that this data could not be collected for nearly 69,000 married or widowed women living there.69
Interestingly, census officials generally agreed that in contrast to so-called ‘illiterates’ who on the whole needed little persuasion to answer the census questionnaire, it was the urban ‘intelligentsia’ who did not always take the process seriously: in Karachi, “a Secretary to the Government had to be visited eight times at his house and office before his statement could be recorded”.70 Addressing a public meeting in federal capital on the first day of the census proper, Census Commissioner Slade called again for wholehearted public cooperation in the weeks ahead.71 After all, as both he and the Administrator of Karachi Hashim Raza emphasised, accurate census results were essential for planning how to build the new state:
Pakistan is still young and impressionable; its future plans and programmes are still on the anvil and its constitution still unshaped. We do not have to repair, renovate or demolish an old construction. We have to raise an altogether new edifice, a task which, although stupendous in itself, is less intricate and more straight [sic]. Census results … are destined to play a significant role in our future development schemes and in shaping a near-utopian society.72
69 Ibid., pp. 192-3. 70 Ibid. This experience chimes with the argument put forward by Ian Talbot on refugee rehabilitation albeit in Indian Punjab, namely that “wealthier refugees were not only more likely to buy into the Punjabi myths of self-reliance, but to have the most to hide in terms of their dealings with the state. Poor migrants, on the contrary [and it might be assumed that illiterate people in early 1950s Karachi were likely to have been low down the economic ladder], had little power or ability to access the state on their own terms” and, hence, would have been more likely to engage with the state, or its representatives, as and when the latter required. See Ian Talbot, ‘Punjabi Refugees’ Rehabilitation and the Indian State: discourses, denials and dissonances’, Modern Asian Studies, Vol. 45, no. 1 (2011), pp. 109-30. 71 Dawn (10 Feb. 1951). 72 A. U. Akhtar, ‘Social Significance of the Census’, Dawn (25 Feb. 1951).
26
On Census Night (28 February-1 March) itself, ‘a hush, ruffled only by gentle knocks and whispers of the enumerators, prevailed’. 73 The streets of Karachi were said to be completely deserted as census workers, permitted to hire rickshaws at government expense, went from house to house with a supply of individual questionnaires. Enumerators were also tasked with tracking down any individuals who might be at danger of being left out of the count because they lacked a permanent lodging place: in effect, the hunt was officially on for the city’s floating homeless population that, thanks to the accommodation shortage, still spent its night on footpaths on in other places not officially on record with the municipal authorities.
What these unfortunates thought about the census remains unknown, but according to the final census report on the city at least, the results of the publicity campaigns that had accompanied the operation there proved—at least officially—to be “excellent”. In its words,
“the public became conscious of the Census and began to realise the importance of the operation; their minds were disembued [sic] of many false notions crediting the Government with ulterior motives”.74 But the reality was that for all the official effort to put a positive gloss on proceedings, the census operation had been subject to criticism. In the run-up to the count, journalists, for instance, took great umbrage at being classed in a similar category to prostitutes, and had to be reassured that “This is not true…Prostitutes are nowhere mentioned in any part of the list. Should they declare themselves as such, they will be entered in a separate category which relates to persons whose occupations are not otherwise classified.
[After all] It must be remembered that in the census persons are classified in accordance with the description they themselves give”.75 In response, Pir Ali Muhammad Rashdi, President of the Pakistan Newspaper Editors Conference (PNEC), made it clear that he did not regard this explanation as satisfactory:
73 Dawn (10 Feb. 1951). 74 Karachi Report, p. 125. 75 Dawn (14 Jan. 1951).
27
The principal point is: is it or is it not a fact that the prostitutes and journalists appeared under the same major head of occupational classification? ... we in this country do not necessarily go by what is done in the West or by the UN. Here we have our own distinct conception of good and bad. There, to be a dancer or a professional singer may mean nothing: here it means a good lot. Therefore I am not impressed by the plea put forward.76
Likewise, during the census period itself, a letter-writer to the Karachi-published newspaper
Dawn berated enumerators for wasting the public’s time by asking apparently meaningless questions:
(to a lady) Are you married? A: “I am single”. Q: Where is your husband? Any children? (To a gent): Are you married? A: “Bachelor”. Q: How many children have you got? Have you any wives? In some cases ladies have been asked “How many husbands have you got?”…The Census Commissioner should have been fully aware of the customs of all the communities residing in Pakistan before instructing the staff…When the enumerators ask “What is your religion?”, why shouldn’t they use their common sense?77
In response, another correspondent suggested pointedly that actions would have spoken more effectively than words, and that the first letter-writer should “have served the Nation better by enrolling himself as an Enumerator instead of acting as an arm-chair critic indulging in mud- slinging”. 78 One unanticipated problem emerged in terms of how to gather data from
Karachi’s reportedly fairly large deaf and dumb community:
Enumerators are finding it difficult to gather particulars about the deaf and dumb who are most homeless and have no neighbours who could tell particulars about them…Enumerators [have] had to leave census slips blank when they come across illiterate deaf and dumb citizens.79
Concluding thoughts
Interim census results were released as early as March 1951. In the main, contemporaries suggested that these agreed pretty closely with earlier estimates made on the basis of the 1941 census and having taking patterns of recent migration into account. All the same, there were
76 Dawn (18 Jan. 1951). 77 Dawn (27 Feb. 1951). 78 Dawn (28 Feb. 1951). 79 Dawn (19 Feb. 1951).
28 suspicions that the official totals had been inflated, most likely for “political reasons”, in certain areas of what had now become Pakistan such as Sindh.80 On the other hand, as the
Census Commissioner later reported, “there is in my mind a grave doubt as to whether such cities as Dacca, Lahore, Rawalpindi, Multan and especially Karachi were 100 per cent enumerated”.81 In the case of Karachi, where Partition generated arguably the most far- reaching demographic fall-out in Pakistan, a recent (mid-1950) estimate of 1,177,226 given for the city’s inhabitants proved be quite close to the first provisional post-census figure of
1,118,000 issued for the Federal Capital Area.82
Moreover, and arguably more significantly, the official census findings confirmed what the federal authorities wanted to hear: that Pakistan’s capital city was considerably more representative of the country-at-large than many contemporaries recognised (or alternatively wanted) it to be. As a later Census Bulletin proudly announced, Karachi’s growth ranked as one of the new country’s “outstanding phenomena”, with the city placed in “the same rank as far as the number of inhabitants are concerned as Athens, Birmingham, Glasgow, Milan,
Melbourne and Montreal”. While “muhajirs from India [now accounted] for more than half the total population of the FCA”, this growth could not be attributed “to their arrival” alone, since “Government Servants, businessmen, skilled workers from all parts of Pakistan have been drawn to the capital and the departure of Hindus and Sikhs (the majority previous to
Partition) has been cancelled out many times”. 83 Karachi, in effect, had moved on dramatically from its pre-1947 days, had no need of those who had left, and now ‘belonged’
80 Press Note, E No. 1061, 20 March 1951, Press Information Department, Government of Pakistan, MSS Eur F 158/628, BL. 81 E.H. Slade, Census of Pakistan 1951, Pakistan Report and Tables, Vol. 1 (Karachi: Manager of Publications, Government of Pakistan, 1955), p. 2 82 Gazette of Pakistan, supplement, n.d., MSS Eur F 158/628, BL. 83 Census Bulletin No. 3, Office of the Census, Government of Pakistan, Ministry of the Interior, Karachi (Sept. 1952), p. 2, MSS Eur F 158/628, BL.
29 to Pakistan as a whole: the implicit conclusion, backed up by census findings, being that neither muhajirs nor Sindhis could claim it as their city alone.
Second, and again specifically in relation to Karachi, census findings provided the federal authorities with information that they needed to realign the city’s internal boundaries albeit retrospectively. By the early 1950s, for instance, doubts had surfaced about the exact boundaries between the Municipal Corporation, the Port Trust and the civil cantonment area.
The census process helped to settle these in a manner that appeared satisfactory to officials at the time, and so played a vital role in re-drawing the map of Karachi. In addition, there were
11 circles in the City Cantonment area (containing over 70,000 inhabitants) which census officials had, conveniently, decided belonged instead to the Corporation. Hence, the census report had published a special table showing population according to sex, religion and the number of muhajirs in them based on redrawn ward boundaries even though these were not actually in existence at the time of the census itself. By an extraordinary order dated 4 June
1952 these were later confirmed as Municipal Corporation wards, based on the population figures produced by the census and the interim divisions on which the census enumeration was conducted.84
1941 1951
Population 450,000 1,137,000
Sindhi spoken as mother tongue (%) 61.2 8.6
Urdu spoken as mother tongue (%) 6.3 50
Hindu population (%) 51 2
Muslim population (%) 42 96
84 Karachi Report, p. 123.
30
Comparison of findings of 1941 and 1951 census findings for Karachi85
A third example of the political significance of the census was the way in which it
exposed a shift in the languages and religions represented locally over the course of the
decade since 1941. Apart from emphasising that well over a third of Karachi’s inhabitants
seem not to have spoken Sindhi as their mother tongue back in 1941—a likely reflection of
the large presence of Gujarati-speakers there, many of whom had left for India by the time of
the 1951 Census, as well as the demographic impact of wartime—and with only very few of
its inhabitants having then declared themselves to be Urdu-speakers, the reconfigured reality
captured in the census reports also highlighted the spectacular decline in the number of
Hindus remaining in the city less than five years after Partition. The ongoing sensitivity
regarding Gujarati during Pakistan’s early years was revealed in a press notice issued before
the census was carried out. In response to anxiety among Gujarati-speakers in Karachi that
their mother-tongue would be omitted from the list of languages printed on the census record
slips, the Census Commissioner had felt it necessary to reassure them that there was no
question of this data not being collected: “the census record slip has a blank space specially
left at the head of the last column for language classification” and enumerators had been
instructed to fill in the relevant information.86 While the 1951 headcount underlined that
Sindh contained the highest percentage of Hindus in its population as compared with other
parts of West Pakistan, again the census’s official findings helped to reinforce the message
that Karachi was now a Muslim, and by implication, a ‘Pakistani’ city.
Between 1941 and 1951, Karachi had added some 691,000 people to its inhabitants,
representing a 158.4 per cent increase. In the decade that followed, its population continued
85 Arif Hasan and Masooma Mohib, Urban Slums Reports: The case of Karachi Pakistan (2003) [http://www.ucl.ac.uk/-4DEB6909-F1924266A6D3F1973C5-200CA/Final- Download/DownloadId-0C55CBDBAA5BB5241B736261F0D48434/4DEB6909-F192-4266 A6D3-F1973C5200CA/dpu-projects/Global_Report/pdfs/Karachi.pdf, accessed 8 May 2016]. 86 Dawn (11 Jan. 1951).
31
to rise, by over 900,000 (80.5 per cent). A 1959 survey to determine the demographic
characteristics of the city in 1959 placed the total at around 1.8 million; by the time of the
1961 census it had topped two million. In terms of percentage increases, however, the rate of
growth slowed over the course of the 1950s, declining to 5.6 per cent per annum between
1959 and 1961 as compared with 6.3 per cent between 1951 and 1956 and 10 per cent
between 1941 and 1951.87 But this pattern was perhaps only to be expected. The influx of
migrants from India slowed down markedly in the second half of the 1950s, in large part
thanks to the strict implementation of immigration rules. 88 In July 1955, the federal
government banned the unauthorised entry of Muslims from India into Pakistan. In October
1959, the Federal Capital Area was shifted to Rawalpindi, which also acted as a brake on
Karachi’s expanding population growth, at least in the short term.89 But while immigration
was declining, in-migrating continued to grow. And just as important were fertility patterns.
As one report explained, “Immigrants from India and people from other parts of Pakistan,
attracted by better work opportunities, have moved into Karachi rapidly and are probably
adopting many new ways of life in the metropolis. But in regards to fertility, they apparently
still think in their old ways and have not changed from their traditional pattern of high fertility
…. administrators and city planners are usually aware and interested in the problem of
migrants but the problem of excess births over deaths attracts less attention … The main
current problem of population growth is therefore no longer a product of net migration but of
procreation”.90
But we should not overlook the fact that, from the perspective of the state, the census
also represented an important rite of passage: as a later press release commented,
87 Sultan S. Hashmi, The People of Karachi: Demographic Characteristics (Karachi: Pakistan Institute of Development Economics, 1965), p. 19. 88 Zamindar, The Long Partition, 89 Ansari, Life After Partition, pp. 197-9. 90 Hashmi, The People of Karachi, p. 125.
32
A population census is an essential requirement of modern administration, amounting as it does to a stock taking of the most valuable asset of the Nation, the people. Censuses have been taken from time immemorial, but in this age of planning, when the people of Pakistan are girding themselves to ameliorate their own condition, the nation needs to know its own strength and quality, the social status of its inhabitants, and their educational, cultural and economic conditions. Information is required regarding languages, occupations and technical ability to man the services and industries. The first Census of Pakistan was therefore planned to measure these resources and assess the ability of the country to shoulder its own burden.91
But, as this article has sought to show, it is also important to acknowledge how far the census operation in Sindh exposed key political and bureaucratic challenges facing the Pakistani state, at both federal and provincial level, in its efforts to establish itself after 1947. The way in which the census was carried out demonstrated the importance that the state attached to the operation as a way of developing and reinforcing its relationship with its new citizens as well as its central role in constructing and knowing whole populations as objects of management and government.92 Thus, official and non-official reactions to the census help us to understand the shaping of state-endorsed as well as more popular understandings of ‘Pakistan’ during a crucial period of political transition there. This conclusion chimes with arguments put forward by Kamal Sidiq in his study of more recent strategies pursued by so-called illegal immigrants to obtain citizenship. Just as Sidiq cautions strongly against regarding citizenship as simply a top-down construction (‘citizenship from above’)—pointing instead to how far individuals can subvert such ordering by accessing rights through alternative, extra-legal channels—so to censuses need to be viewed as bottom-up exercises or at least occasions when
91 ‘Pakistan Day: Fourth Anniversary Series’, Press Information Department, Article No. 27/51, ‘The First Census of Pakistan – 1951’, MSS Eur F 158/628, BL. 92 For an exploration of the biopolitical significance of census operations in another part of the world—in effect, the use of social statistics as modern phenomena for governing populations and nation-building—see Jan Busse, ‘The Biopolitics of Statistics and Census in Palestine’, International Political Sociology, Vol. 9, no. 1 (March 2015), pp. 70-89. See also for a European perspective on the subject, Evelyn Ruppert, ‘Making Populations: From Censuses to Metrics’, in Sichtbarkeitsregime: Überwachung, Sicherheit Und Privatheit Im 21. Jahrhundert, eds. Leon Hempel, Suzanne Krasmann and Ulrich Bröckling (Wiesbaden: VS Verlag, 2010), pp. 157–73 [http://oro.open.ac.uk/26069/, accessed 8 May 2016].
33 those being counted may or may not choose to cooperate with the state and its ‘everyday’ representatives.93
Thus, as this article has highlighted, contextualising the 1951 census allows us to gauge the significance of number-counting from the perspective of early state-building efforts in Pakistan, and what these meant at the local level, in this case Sindh and in particular the city of Karachi located there. In time, the legacies of this first census—above all, the officially-verified increase in the number of declared Urdu speakers—came to bear upon federal and provincial government policies, with the census figures themselves also increasingly appropriated or undermined by the various communities now living in this part of
Pakistan. By the early 1970s, these pressures came to a head, fuelling growing tension between Sindhi and muhajir groups in provincial political life. Sindhi interests, it seemed, now clashed very openly with those of former refugees from India. First, language became a clear-cut bone of contention when the provincial Pakistan People’s Party ministry, led by
Mumtaz Ali Bhutto, introduced the Teaching, Promotion and Use of Sindhi Language Bill in
July 1972, making Sindhi a compulsory subject at school level and also raising the prospect of it taking the place of Urdu in local government offices and courts. Language riots followed.
Then, to further complicate the picture, a new federal quota system for government employment was introduced in 1973, which split Sindh’s share of just under 20% of federal jobs into two sections—rural (11.4%) and urban (7.6%)—with Sindhis largely falling into the first category and muhajirs the second, a move that in practice threatened the latter’s access to
93 Kamal Sidiq, Paper Citizens: how illegal immigrants acquire citizenship in developing countries (Oxford: Oxford University Press, 2008).
34 federal jobs, and formed part of the backdrop to the emergence of the Muhajir Quami Mahaz
(MQM) as a key political force in Sindh’s cities and towns in the 1980s.94
Finally, as Akhil Gupta has emphasised in Red Tape—his hugely-influential anthropological study of present-day bureaucratic practices in India—we must remember that
“states are not simply functional bureaucratic apparatuses but powerful sites of symbolic and cultural production that are themselves always culturally represented and understood in particular ways”. 95 Indeed, exploring how Pakistan’s first census was conducted in Sindh reinforces the importance of thinking about the ways in which statistical data is collected and the mechanisms that determine the effectiveness of this biopolitical exercise. Viewed from the perspective of twenty-first century Pakistan and bearing in mind the combination of political and logistical problems that has hampered census operations in the region for nearly twenty years—Pakistan’s long-awaited latest census promised for 2016 having (at time of writing) yet to materialise—number-counting, together with the politics of number-crunching, continues to represent a matter of logistical and numerical anxiety for the Pakistani state and its citizens alike.96
94 For a careful contextualisation of the rise of the MQM and other developments concerning muhajir politics, see Laurent Gayer, Karachi. Ordered Disorder and the Struggle for the City (London: Hurst, 2014). See also Matthew A. Cook, ‘Introduction: Mohajirs, Sindh and Pakistan’, in My Life's Journey, The Early Years (1966–1988) by Altaf Hussain (Karachi, Oxford University Press, 2011). 95 Bob Jessop, “Narrating the Future of the National Economy and the National State: remarks on remapping regulation and reinventing governance”, in State/Culture: State Formation after the Cultural Turn, ed. G. Steinmetz (Ithaca: Cornell University Press, 1999), pp. 379-405, cited in Gupta, Red Tape, pp. 42-43. 96 ‘Pakistan to hold population census in March 2016, with military support’, Dawn (18 March 2015) [http://www.dawn.com/news/1170388, accessed 8 May 2016]; ‘CCI decides to postpone population census’, 29 Feb. 2016 [https://www.geo.tv/latest/101596-CCI-decides-to-postpone- population-census, accessed 8 May 2016].